Relay minimum contact current/voltage

[ShayneC]

Hi,
I'm designing a voltmeter circuit that will be mostly measuring voltages below 5V ac and dc but it does require the capability to measure up to 250V very occasionally.
So I have designed an auto range system using relays to switch between the different ranges. But I need the device to have very good battery life and as it will spend most of its time on the 5V range I am using latching relays so I don't have to waste battery power switching them often. I'm also running the system from parallel lithium ion batteries for a nominal battery voltage of around 3.7V. I don't want the noise from a buck convertor and only want LDO's to get the correct voltage for the ESP32 I'm running the system from. I'm struggling to find a suitable relay though. There aren't many 3V relays to choose from. The relay in this link

https://www.mouser.com/datasheet/2/447/KEM_R7002_EC2_EE2-3316920.pdf

meets my requirements apart from one spec. The minimum contact current. It says it needs 10mV, 10uA. Does this mean it needs at least 10mV and 10uA. Or one or the other. If it's one or the other I'll be OK as the voltage will be above 10mV. But the input impedance is really high so the current will be really low.

If it's both I have an idea. If I use another relay to link the battery directly, with a current limiting resistor, to the circuit and push a few mA of current through the relays for a few ms every time the relay gets switched, before I use the circuit to take measurements, will this keep the contacts clean and compensate for the really low current I will usually be using through the relays? Connecting the battery to the circuit I'm measuring will not be detrimental to it.

Thanks for your help.

[Postal2]

... I am using latching relays ...

OK, it's right.

.... There aren't many 3V relays to choose from. ...

You need a simple pump L + fet. C accumulate voltage and discharge to relay.

... Does this mean it needs at least 10mV and 10uA. ...

You need gold contacts to measure weak signals. You will have jumpy measurements on the contacts you specified.

[SeanB]

Yes for driving the relays you want some 5V or 12V relays, and simply put in a small switched boost converter, that you enable a half second or so before the switch, to provide a higher voltage bus for them. For low current contacts you definitely want gold coated contacts, which will hold up for a while. With the relays best as well to use a double pole relay, and use the second contact to provide feedback for contact position, so that you can have the system able to detect range on power up.

[Postal2]

.... use the second contact to provide feedback for contact position ...

This is an unwise waste of gold and money, don't do it.

[exmadscientist]

Be careful of latching relays, the state they come up in is random. (OK, it's determined by whatever happened to things mechanically -- so, random.) This can be bad. I tend to avoid them for that reason, but sometimes you have to deal with it.

The minimum current is required to keep the contacts clean. It applies in the CLOSED or closing condition so of course voltage across a dead short will be small. This can get kind of weird so I'd suggest you search it (relevant term: wetting current) rather than me trying to explain from a phone.

If you are worried about low level signals you want fancy contacts. Pd is said to be the best. I don't know about best but I've used a lot of Pd or Ru contact relays (Panasonic AGQ, Fujitsu FCL FTR-B3, TE Axicom IM) and they seem to do very well. 3V coils and below are widely offered but I don't know about stocked. Somehow, manufacturers actually managed to standardize these things, so they're interchangeable! For once.

Since you want to handle 250V, I would look at the TE Axicom IM series. It looks like it would do everything you want, if there's stock of an acceptable part number.

Otherwise, in light of 250V being a very rare rating for a signal relay, consider whether (1) you need to switch it at all (no switching means you're on the much more forgiving standoff rating table) and (2) whether you can bring in another series switch (maybe semiconductor?) to take the voltage stress off the relay.

[Doctorandus_P]

From what I know of your circuit, I would not use a relay at all. Voltage meters are high impedance circuits, and relay contacts are not great for very low currents. A relay is also not needed at all. Just start with a high resistance resistor (Capable of handling your input voltage!). and after that, add some clamping diodes. Whenever your input voltage is "too high", you can switch in a secondary resistor to create a voltage divider. This can be a simple CMOS switch such as an 4066, or one of those ADG things if you need them and can afford it. You do need to think a bit about your circuit to make sure the voltages on the switch is always within their operating range.

[Karel]

I used these relays in my 6.5 digit DVM project and they work just fine (contacts are gold plated):

Bistable, 2 coil, 5VDC.:

Manufacturer: AXICOM - TE CONNECTIVITY
Part number: V23079B1201B301

https://uk.farnell.com/axicom-te-connectivity/v23079b1201b301/relay-signal-dpdt-250vac-220vdc/dp/4219995

For 3VDC: part number V23079B1208B301

https://www.digikey.com/en/products/detail/te-connectivity-potter-brumfield-relays/V23079B1208B301/2398081

(there are also versions with one coil)

[ShayneC]

Thanks for all your suggestions.
I'm going to build my prototype board utilizing a few different options so I can test them all out and see which performs the best.
Another idea I also just came up with is to just get an ADC with an extra channel and just do away with all the relays and plug two different voltage ranges into the two different channels. Then just take my readings from the higher channel when the first has maxed out. I should be able to get away with just two voltage ranges for my product.